THEEQUILIBRIUM ROH -4- HC1 $ RC1
July, 1941
[CONTRIBUTION FROM
THE
+ HzO
1873
CHEMISTRY LABORATORY OF THE UNIVERSITY OF MICHIGAN ]
Equilibrium in Chloride Formation from Mono-, Di- and Triarylcarbinols BY J. 0. HALFORD AND EUGENE B. REID’ X quantitative study of the equilibrium in the reaction ROH
+ HCl
RCl
+ HzO
(1)
has been undertaken for a series of compounds in which R is a mono-, di- or triarylmethyl radical. A list of the compounds selected for study appears in Table I. For these substances it has been found that equilibrium can be attained a t 60’ in a reasonable length of time, and that the reaction proceeds according to Eq. 1 without interference from the formation of other products. The first experimental step follows a procedure used by Halford2 in a study of the hydrolysis of triphenylchloromethane and triphenylbromomethane. A dilute benzene solution of the carbinol and chloride is brought to equilibrium with aqueous hydrochloric acid and the phases are separately analyzed. An equilibrium constant K(1) is evaluated in accordance with Eq. 2, in
which C represents concentration in benzene and p is the partial pressure taken from “International Critical Tables. ” 3 In the present work, for the reactions involving the benzyl and p-xylyl alcohols and the diphenyland di-p-tolylcarbinols, the results have been extended t o the vapor phase with the aid of partial pressures in benzene solution, to give vapor phase equilibrium constants which should not be in error by more than 15y0. The corresponding extensions for the remaining diary1 compounds and for the non-volatile triaryl compounds are based upon estimates of the partial pressures and are therefore subject to a somewhat larger uncertainty. Solutions of benzyl chloride, triphenylchloromethane and triphenylcarbinol in benzene are nearly perfect and permit the use of the relation p = poN as a first approximation in obtaining the vapor phase constant. It is assumed that all of the chlorides and tri-p-tolylcarbinol will show similar behavior. On the other hand, the relation (1) From a dissertation submitted by Eugene B. Reid in partial fulfillment of the requirements for the degree of Doctor of Philosophy i n the University of Michigan, 1939. (0) J. 0. H a l f o r d , T ~ m JOURNAL, 53, 105 (1931). (3) “ I . C. T.,” Vol. 111, p. 216.
between partial pressure and mole fraction for the mono- and diarylcarbinols in benzene is more complicated, with the former showing the greater deviations from the perfect solution equation. We find that the partial pressure of the solvent is adequately expressed for the benzyl alcohol solutions by the equation p1 = PYX1 where XI = 1 - Xz= ANl/(ANl Nz). It follows directly that A = (:Vz/Nl) (pi/ [p: - pi]), and in conformity with the Duhem-Margulis equation, that pz = pi Xi’”. Since the solutions of p-xylyl alcohol and the diarylcarbinols in benzene are more nearly perfect than those of benzyl alcohol, it is safe to assume that the above relations may be used for all the carbinols. The vapor phase constant may now be expressed as HzO = K ( g ) = PRCIPB20 = Pic1 -NRCl -P__ PROHPHC~ PEC,
+
~
It follows from the definition of X that a t low concentrations XRoH = NROH/A. Consequently
In deriving Eq. 4 the concentration ratio of Eq. 2 has been replaced by the equivalent ratio of mole fractions. The factor by which the liquid phase constant must be multiplied in order to give K(g) involves the vapor pressures of the pure liquid carbinol and chloride and the quantity A , which latter depends upon the partial pressure of benzene over solutions of the carbinols. Although this treatment of the carbinol solutions provides a correction for association, its success cannot be explained in terms of a simple picture of the association equilibrium. If, for example, only a monomer and dimer were involved, and each acted as a perfect solute, the correction factor should increase with dilution. The observed situation suggests a stable dimer in equilibrium with a monomeric form of high fugacity, Table I presents a summary of the liquid phase constants defined by Eq. 2. The vapor phase constants are included, for comparison, as the
18'74
J. 0. HALFORD AND EUGENE B. REID
~ ~ Q U I l . l B R I t i IN M THE
I
